Effects of changes in land-use and natural disasters on social-ecological resilience and vulnerabilities in coastal Bangladesh
Md Modasser Hossain Khan
Department of Environment and Development Studies, NORAGRIC Master Thesis 30 credits 2012
i
The Department of International Environment and Development Studies, Noragric, is the international gateway for the Norwegian University of Life Sciences (UMB). Eight departments, associated research institutions and the Norwegian College of Veterinary Medicine in Oslo.
Established in 1986, Noragric’s contribution to international development lies in the interface between research, education (Bachelor, Master and PhD programmes) and assignments.
The Noragric Master theses are the final theses submitted by students in order to fulfil the requirements under the Noragric Master programme “International Environmental Studies”,
“Development Studies” and other Master programmes.
The findings in this thesis do not necessarily reflect the views of Noragric. Extracts from this publication may only be reproduced after prior consultation with the author and on condition that the source is indicated. For rights of reproduction or translation contact Noragric.
©Md Modasser Hossain Khan, December 2011 [email protected]
Noragric
Department of International Environment and Development Studies P.O. Box 5003
N-1432 Ås Norway
Tel.: +47 64 96 52 00 Fax: +47 64 96 52 01
Internet: http://www.umb.no/noragric
ii
DECLARATION
I, Md Modasser Hossain Khan, declare that this thesis is a result of my research investigations and findings. Sources of information other than my own have been acknowledged and a reference list has been appended. This work has not been previously submitted to any other university for award of any type of academic degree.
Signature………..
Date………
iii
DEDICATION
I dedicate this thesis to my parents who inspired me throughout the data collection and writing period and encouraged me until the end, to complete the hard work successfully.
Without them, this would not have been possible.
iv
ACKNOWLEDGEMENT
First and foremost, I would like to extend my sincere thanks and greatest appreciation to my supervisor, Prof. Ian Bryceson, for his valuable guidance, advice and support. He has been extremely helpful throughout the research period. Without his encouragement and guidance, this thesis could not have materialized. I am also thankful to Ass. Prof. Siri Eriksen, for her thoughtful advice and guidance, which helped me a lot throughout the research period. Special thanks to my study coordinator Ingunn Bohmann for her kind support and help during the whole research period.
Special thanks also go to Dr. Ubydul Haque of IMT department, for his guidance and for helping me throughout the image analysis procedure. I would also like to thank Hafizur Rahman, PhD student of IPM department for his support and help during the writing period.
The work leading to this study was supported by a research grant from Noragric, and this financial help is greatly acknowledged. Without the financial support, it would not be possible for me to conduct the fieldwork. I would like to thank all the members of Noragric for their cooperation’s.
v ABSTRACT
Natural disasters and land-use change are major concerns all over the world, and if these two concerns exist together in a coastal area, then the consequences for people and the environment may be severe. This study investigated the changes in land-use in the past 10 years in Shyamnagar Upazila of the south-west coastal area of Bangladesh. The drivers of land-use change and the occurrence of disasters were explored in relation to their effects on social and ecological systems. Satellite images were analyzed to detect changes in land-cover in the last 13 years. Three areas were selected for on-the-ground data collection. Household surveys were conducted to discover the type, level and effects of disasters. Focus Group Discussions and personal interviews were also conducted to explore the drivers behind changes in land-use.
Probability regression analysis was performed to assess the relationship between various disasters with overall income, agricultural production and outward migration. Results from image analysis showed an overall 21 percent increase in shrimp culture ponds in the past 13 years. Agricultural land and forest resources decreased by 48 and 3 percent, respectively, while barren and built up areas increased by 71 percent. Analysis of household data showed that cyclones and tidal floods had significant effects on income, agricultural production and migration. Social, economic and political factors combined with natural causes were found to be the main drivers behind land-use changes. These empirical findings suggest that social and ecological resilience was reduced and vulnerabilities increased in this part of coastal Bangladesh for these reasons between 1999 and 2012.
vi
TABLE OF CONTENTS
1. Introduction ... 1
1.1 Background ... 1
1.2 Research Questions ... 1
1.3 Objectives ... 2
1.4 Hypotheses of the study ... 2
1.5 Conceptual Framework ... 2
2. Review of literature... 3
2.1 Land-use and land-cover changes ... 3
2.2 Detection of land-use and land-cover changes using satellite imagery ... 3
2.3 Land-use practices in coastal areas of Bangladesh ... 4
2.4 Disasters in coastal areas of Bangladesh ... 5
2.5 Vulnerability and resilience in terms of disasters and land-use changes ... 5
3. Methods... 8
3.1 Selection of the study area ... 8
3.2 Description of the study area ... 8
3.3 Satellite Image analysis ... 9
3.4 Data collection and analysis ... 10
3.5 Limitations ... 10
4. Results ... 11
4.1 Land-cover change detection ... 11
4.1.1 Change detection Statistics ... 16
4.2 General characteristics of respondents ... 16
4.3 Changes in land-use and their causes ... 17
4.4 Factors responsible behind major land-use changes ... 18
4.5 Disasters in the study area and their effects according to respondents ... 19
4.5.1 Salinity level of the study area ... 19
vii
4.5.2 Type and level of disasters on the study area ... 19
4.5.3 Impacts of disasters on agricultural production, income and migration ... 21
4.5.4 Consequential effects of disasters ... 23
4.5.5 Diseases outbreak after disasters ... 24
4.6 Migration in the study area and their causes ... 24
4.7 Changes in tree and fish species ... 25
5. Discussion ... 27
5.1 Changes in land-use and land-cover in the study area from 1999 to 2012 ... 27
5.2 Effects of land-use changes ... 28
5.3 Drivers behind major land-use changes ... 29
5.4 Effects of disasters on the study area ... 29
5.5 Causes of migration in the study area ... 30
5.6 Resilience and vulnerabilities of the study area ... 31
5.6.1 Social-ecological resilience and thresholds ... 31
5.6.2 Vulnerabilities of social-ecological system ... 32
6. Conclusions ... 34
7. References ... 35
8. Appendices ... 39
8.1 Appendix 8: Household questionnaire ... 39
8.2 Appendix 2: FGD Questionnaire ... 45
8.3 Appendix 3: List of occurrences of natural disasters on last 13 years ... 48
8.4 Appendix 4: Cyclone and tidal surge risk areas of Bangladesh ... 49
viii
LIST OF TABLES
Table 4.1: Change detection Statistics from year 1999 to 2012………16
Table 4.2: Baseline characteristics of the study population (N indicates number of respondents)………...17
Table 4.3: Changes in local practices and their causes according to respondents……….18
Table 4.4: Land status in study area before shrimp cultivation……….18
Table 4.5: Factors responsible for land-use changes……….19
Table 4.6: Level of Salinity in study area……….19
Table 4.7: Type of disasters in the study area………20
Table 4.8: Effect of disasters on agricultural production………...21
Table 4.9: Effect of disasters on income of the respondents……….22
Table 4.10: Effect of disasters on outwards migration………..22
Table 4.11: Disasters and their consequential effects on the study area………..23
Table 4.12: Diseases outbreak after latest disaster (Cyclone Aila 2009)………..24
Table 4.13: Outward migration in study area………25
Table 4.14: Causes of outward migration in study area……….25
Table 4.15: Changes in tree and fish species from 1999 to 2012 (according to respondents) in study area………...26
LIST OF FIGURES Figure 3.1: Map of Shyamnagar upazila of Bangladesh (1, 2 and 3 indicates those areas where field survey was conducted)……….8
Figure 3.2: Image analysis procedure………..9
Figure 4.1: False color images of Shyamnagar and its surrounding areas from year 1999 and 2012………11
Figure 4.2: Classified map of the study area from the year 1999………..12
Figure 4.3: Classified map of the study area from the year 2012………..12
Figure 4.4: Change map bare land and built-up areas from1999-2012……….13
Figure 4.5: Water body change map 1999-2012………14
Figure 4.6: Agriculture land change map from 1999-2012………...15
Figure 4.7: Vegetation change map 1999-2012……….15
Figure 5.1: How disasters and LULCC increasing Vulnerability (Adapted from O’Brien et al. 2007)………..33
ix
LIST OF ACRONYMS
NDRI Natural Disaster Risk Index
MWR Ministry of Water Resources
BBS Bangladesh Bureau of Statistics
UNISDR United Nations International Strategy for Disaster Reduction UNEP United Nations Environment Programme
UN-HABITAT United Nations Human Settlements Programme USGS United States Geological Survey
ENVI ENvironment for Visualizing Images
IDL Interactive Data Language
TM Thematic Mapper
ETM+ Enhanced Thematic Mapper Plus
DN Digital Number
RMSE Root Mean Squire Error
FGD Focus Group Discussion
SAAO Sub-Assistant Agricultural Officer
RA Resilience Alliance
HYV High Yielding Varity
1 1. INTRODUCTION
1.1 Background
Bangladesh is a country of low-lying deltaic floodplain with a coastline of about 710km (MWR 2005). The coastal areas of the country are exposed to various natural hazards such as cyclones, storm surges, sea level rise, floods and droughts due to its vulnerable topography and geographical location. The Natural Disasters Risk Index (NDRI 2010) ranked Bangladesh as the country most vulnerable to natural disasters. Various anthropogenic activities such as pollution, deforestation and water logging have also adversely affected the country. Both natural and anthropogenic activities together intensify damage to the ecosystem and hamper the economy, livelihoods and development of the coastal areas of Bangladesh (MWR 2005). In the last 50 years, many major changes have been made in land-use in the coastal areas of Bangladesh, due to interests competing over land and various natural resources. Major changes have happened in the agricultural sector due to salinisation and embankments (Islam 2006) and several natural disasters (Kumar et al. 2010).
According to the Millennium Ecosystem Assessment, modification of land-use by several anthropogenic activities affects social-ecological systems and therefore, loss of biodiversity (Sarukhán & Whyte 2005). Over time, changes in land-use have occurred and, as a result, social- ecological resilience of coastal areas has been affected (Atwell et al. 2008).The purpose of this study was to investigate changes in land-use practices in coastal areas of Bangladesh over the past 13 years and to explore how social and ecological resilience has been affected. Remote sensing technology was used to detect land-use changes in the study area. Remote sensing technology has the capability to extract spatial data by analyzing satellite images (Tayyebi et al.
2008).
1.2 Research Questions
What changes have occurred in the agriculture, fisheries and forestry sectors in Shaymnagar Upazila (sub-district) of Bangladesh over the past 13 years? What are the main drivers of these changes?
2 1.3 Objectives
To detect how land-use has changed over the past 13 years in the study area by using satellite images;
To investigate the main drivers of land-use changes by conducting on-the-ground fieldwork;
To collect data regarding occurrences of natural disasters and their impacts over the past 13 years on the coastal area of Bangladesh.
1.4 Hypotheses of the study
Vegetation and agricultural land are being converted into aquaculture ponds;
Vegetation is being converted into agricultural land;
Migrations are taking place due to occurrences of natural disasters.
1.5 Conceptual Framework
Resilience is adopted as an analytical approach for understanding linked social-ecological systems and their changing processes (Berkes & Folke 2000). The concept of resilience in terms of disasters is defined as “the ability of a system, community or society exposed to hazards to resist, absorb, accommodate to and recover from the effects of a hazard in a timely and efficient manner through the preservation and restoration of its essential basic structures and functions”
(UNISDR 2004). Livelihood strategies of communities are employed according to their ability to cope with risks that come from shocks by both natural and human factors, and by their relative vulnerability to disasters. Vulnerability is a concept that links the relationship that people have with their environment to social forces and institutions and the cultural values that sustain and contest them. “The concept of vulnerability expresses the multidimensional effects of disasters by focusing attention on the totality of relationships in a given social situation which constitute a condition which, in combination with environmental forces, produces a disaster” (Bankoff et al.
2004).
This study utilized both resilience and vulnerability concepts to gain an overview of the situation in the study area. These concepts helped to analyze and understand the changes that occurred in land-use during the last 13 years due to various factors (shocks) and the present condition (responses) of the social-ecological systems.
3 2. REVIEW OF LITERATURE
2.1 Land-use and land-cover changes
Land-use has been seen as a product of interactions between a society's cultural background, skills and physical needs on the one hand, and the natural potential of land on the other hand (Nagamani & Ramachandran 2003). According to the United Nations Food and Agriculture Organization “Land use is characterized by the arrangements, activities and inputs people undertake in a certain land cover type” (FAO 2000). According to this definition, land-use reflects human activities such as the use of the land in industrial zones, residential zones and agricultural fields. Quentin and coworkers define land-use change as a “type of change (physical, chemical or biological) that relates to management like conversion from grazing to crop land, irrigation installation, improvement of drainage systems, building dams, land degradation and pollution, plantation, removal of vegetation, weeds and exotic species increase and conversion to non-agricultural uses” (Quentin et al. 2006). Conversion and modification are two broad categories of land-use and land-cover changes. Conversion refers to changes from one category of land-use to another. Modification includes changing the components of any existing land-use category (Baulies & Szejwach 1997). Significant changes in land-use and land-cover are occurring worldwide (Xing et al. 2006). Land-use dynamics could play a major role in driving changes in the global environment during the next decades (Baulies & Szejwach 1997)
2.2 Detection of land-use and land-cover changes using satellite imagery
Information regarding land-use change is necessary for many applications, such as the monitoring, management and planning of natural resources. It is also important for applications like agriculture, hydrology, forestry, and ecology (Ellis 2010). It has become a central component of current strategies for managing natural resources and monitoring environmental changes. Studies of land-use change have become very important because of the rapid development of land-use mapping (Stathakis et al. 2010). Use of satellite imagery for change detection is a convenient approach to obtaining accurate information on land-use change, because change detection is a major application in digital image processing. Karanja defines change detection as “a technique that is used to highlight conversion of land from one use to another within a given time frame” (Karanja 2002). Land-cover can be determined by analyzing satellite and aerial imagery. Land-use cannot be determined from satellite imagery. Information
4
obtain from land-cover maps help investigators best understand the current landscape. Change detection can be done digitally using satellite images if the images are classified and compared for changes. The overall objective of classification is to categorize all pixels of an image into land-use classes. There are several methods to obtain information about the earth but the most efficient and economical method is through satellite data (Stathakis et al. 2010).
There are two main methods of image classification: unsupervised and supervised classification (Hasmadi et al. 2009). Simple methods such as unsupervised classification of images for change detection suffer from poor accuracy (Hegde 2003). Unsupervised classification is often inappropriate in geographical and remote-sensing sciences (Hegde 2003). To avoid these problems, this study will use supervised classification.
2.3 Land-use practices in coastal areas of Bangladesh
Land-use and land-cover changes in coastal areas of Bangladesh are diverse. Land in these areas is used intensively for agriculture, settlements, forests, shrimp ponds, water bodies and fisheries, salt production, industrial and infrastructure developments and tourism (Islam 2006). Moreover, construction of dams or polders in the coastal areas is considered as one of the main land-use activities that has changed the stability of the area. The sedimentation process in the coastal areas was greatly hampered by the implementation of the coastal embankment project in the 1960s. A total of 97 polders were constructed, of which 37 were in the southwest coastal region. The primary purpose of the project was to protect the agricultural lands from yearly inundation by floods and to free the wetlands from saline water. These polders restricted the river water from entering the wetlands and, as a result, tidal sediments increased the height of these river beds instead of the wetlands. Most river beds became higher than the adjacent wetlands, creating water logging in the surrounding areas. The water logging problem has become more severe and extensive due to saline water intrusion into the surrounding areas at high tides.
The dense population of the country results in a high human-land ratio of about 0.124 hectare per person (BBS 2001). The population of coastal Bangladesh is expected to increase to 43.9 million by2015, compared with 36.8 million in 2001 (Islam 2006). Almost 54 percent of the population is functionally landless in coastal Bangladesh. Among these, 30 percent are absolutely landless (Islam 2006). Almost 71 percent of the 6.85 million households of coastal areas are small-scale farmers, fisher folk, agricultural laborers and the urban poor (Ahmad 2004). This has a
5
significant influence on the quality of land and also on land-use. Almost 50 percent of the coastal lands have lost their effective usefulness due to different degrees of saline inundation (Islam 2006).
2.4 Disasters in coastal areas of Bangladesh
Natural disasters are defined as “a serious disruption of the functioning of society, causing widespread human, material or environmental losses which exceed the capacity of the affected society to cope using only its own resources” (UNISDR 2004). The coastal areas of Bangladesh are disaster prone because of their geographical location, land characteristics and funnel-shaped characteristics (UNEP 2001). The country has been impacted by a range of natural disasters throughout its history, including cyclones, tidal surges, floods, droughts, tornadoes and river bank erosion. Bangladesh remains one of the worst sufferers of cyclonic casualties in the world.
Nearly 70 major cyclones have hit the coastal areas of Bangladesh during the last 200 years (Mallick et al. 2009). Almost 900,000 people have died in last 35 years due to catastrophic cyclones (Islam & Ahmed 2001). The number and severity of cyclones in Bangladesh and the associated human mortalities have varied greatly during the past 50 years. For example, the deadliest cyclone occurred in 1970 with a wind speed of 222 kilometers per hour and a surge height of 5.5 meters, causing almost 500,000 deaths, and the most recent severe cyclone of 2007, with a wind speed up to 240 kilometers per hour and a surge height of 5 meters, caused 4,234 deaths (Haque et al. 2011). In addition, many other hazards affect the country, such as earthquake, tsunami, high arsenic levels in ground water, water logging and salinity intrusion.
Among these, cyclonic surges, tornadoes and droughts are caused by natural processes. On the other hand, river bank erosion, coastline erosion, salinity intrusion, water logging and decreases in groundwater levels are caused by a combination of both natural and anthropogenic activities.
Most man-made coastal disasters have been further accelerated by the upstream diversion of the Ganges in India. The flood situation all over the country, including low lying coastal areas, has accelerated due to damming, polders and the Farakka Barrage of the Ganges in India (Khalequzzaman 1989).
2.5 Vulnerability and resilience in terms of disasters and land-use changes
The effects of natural disasters are shaped by the nature of the vulnerability surrounding particular land governance types (UN-HABITAT 2010). Land and land-use maybe affected
6
differently by various types of disaster. Large areas maybe left uninhabitable by long-term inundation through flood, storm surge and tsunamis. Cyclones, tornados and other big wind events do not necessarily have a large physical effect on land but may cause destruction of houses and resources and displacement of a large number of people (UN-HABITAT 2010).Wind and tidal action have immediate or short-term effects, but the effect of erosion and saltwater intrusion may be long-term, varying from months to years (Wisner et al. 2004). Disasters such as floods, drought and tropical cyclones can have potentially severe biological and epidemiological consequences as secondary effects (Wisner et al. 2004).
Different types of natural disaster and social and political disturbances have various effects on the social-ecological system. The social-ecological system is an ecological system that is intricately linked with and affected by one or more social systems (Anderies et al. 2004). The social-ecological system is a process related to the idea that human action and society are linked with nature, and that it is thus unreasonable to make any distinction between natural and social systems (Adger 2006). Vulnerability covers those characteristics of a social system that influences the ability to predict, deal with, resist and recover from the impact of a natural hazard (Wisner et al. 2004). Vulnerability depends on both the sensitivity and resilience of any system that comes in contact with the hazard (Turner et al. 2003). Sensitivity is the degree to which a system is affected, and exposure refers to the inventory of elements in an area in which hazard events may occur (UNISDR 2004). Any system can therefore be highly resilient, but if its sensitivity and exposure are high then it is vulnerable (Miller et al. 2010). Exposed ecosystems with low resilience might still be able to maintain functions and generate resources and other necessary ecosystem services. If the ecosystem faces disturbances similar to high intensity events, then it might shift to a less desirable state and might exceed critical thresholds.
Thresholds are the points at which a relatively small change in external conditions causes a rapid change in the structure or function of the ecosystem (Walker & Meyers 2004) As social- ecological systems are linked, changes in one system may lead to impacts on another, resulting in shifts in both ecosystems and social systems at many different spatial and temporal scales.
Social and ecological development maybe significantly interrupted, and livelihood options may be reduced, leading to environmental refugees, as a result of the impact on the ecosystem’s life- support (Folke et al. 2002). The Stockholm Resilience Centre suggests that “human adaptation has caused loss of resilience and pushed many ecosystems close to thresholds.”
7
(http://www.stockholmresilience.org/research/researchthemes/adaptivegovernancenetworksandle arning/multilevelgovernance.4.aeea46911a3127427980007069.html). Loss of resilience implies that a small disturbance that could previously be absorbed and generate renewal may instead move a community over a threshold into other stability domains. Such change has been referred to as ‘regime shifts’ (Scheffer et al. 2001) and a system with lower resilience tends to shift more easily than a system with ample resilience. These shifts occur in nature but tend to be exacerbated by humans (Scheffer et al. 2001).
Hazards in coastal areas may become disasters due to loss of resilience created by either environmental changes or human actions (Adger et al. 2005). For example, in 1992 Florida was affected by a Category 5 storm, Hurricane Andrew, which destroyed $26.5 billion of property and caused the death of 23 people. However, in 1991 Bangladesh was struck by a same category tropical cyclone and as a result 100,000 people lost their lives and a million others lost their houses and properties due to widespread flooding (Adger et al. 2005). This may be due to Florida’s strong social resilience which led them to deal with the crisis while on the other hand Bangladesh, with weak resilience and social vulnerability, was susceptible to the large-scale destruction of human life.
In countries with a large, dense population and considerable inequality in land-use and income, rural concentrations in high-risk coastal areas can vary greatly. In these cases, local poor people continue to remove protective vegetation, destroy buffer zones and increase their vulnerability to disasters because of the extreme pressure on the land (Wisner et al. 2004). In various natural disasters, the vulnerability of human populations is based on where they have settled, their use of natural resources and the resources they have with which to cope (Adger 2006). A vulnerable society may find it hard to rebuild their livelihoods after a disaster, and subsequently become more vulnerable to the next hazard event (Wisner et al. 2004).
8 3. METHODS
3.1 Selection of the study area
The study area is Shyamnagar Upazila in the Sathkhira district of Bangladesh. Natural disasters such as cyclones, storm surges, tidal floods, saline water intrusion and water logging are prominent features of this area. In addition, land-use and land-cover change is another major concern. The human-induced shrimp farming initiates salinity that seriously affects agricultural production and makes the region vulnerable to unsafe drinking water.
3.2 Description of the study area
Shyamnagar (Figure 3.1) is the largest Upazila, of Satkhira district. This Upazila occupies an area of 1968.24 km2, including 1,622.65 km2offorest area. It consists of 13 unions, and the total population of this Upazila is 313,789 (BBS 2001). Household income predominantly depends on agriculture. Almost 65 percent of the total population is involved in agriculture and 38 percent depend on cropping, livestock, forestry and fisheries as their main source of income, whilst 27 percent derive their income from selling agricultural labor. The total cultivable land is 38,552 hectares, with 6,258 hectares of fallow land. Among the total population, 19 percent are landless, 30 percent landed but small, 28 percent are marginal, 16.5 percent intermediate and 6.5 percent are wealthy (BBS 2001).
Figure 3.1: Map of Shyamnagar upazila of Bangladesh (1, 2 and 3 indicates those areas where field survey was conducted)
9 3.3 Satellite Image analysis
Satellite images have been analyzed to answer the research questions and fulfill the objectives of the study. Images have been obtained from the United States Geological Survey (USGS). Two sets of satellite images, dated1999 and 2012, have been used for the analysis. Images obtained in 2012 had a gap mask due to the Scan Line Connector of Landsat ETM+ being offline (SLC off).
The mosaicking method was applied to fill those gaps using ENVI\IDL TM. For minimizing the differences in the Digital Number (DN) value of each pixel, radiometric correction procedure was applied; then Ground Control Points were collected to register these two images. The root mean square error (RMSE) was 0.3. Supervised classification with a maximum likelihood algorithm was applied for image classification. Later, two classified images were used for change detection. Figure 3.2 represents the overall steps used for the image analysis procedure.
Figure 3.2: Image analysis procedure Output Map using GIS
Trial and Error
Collection of Ground Control Points Landsat Image Collection
Image Calibration
Image Registration
Layer Stacking
Subsetting\Clipping
Image Classification
Landsat 5 TM 1999 Landsat 7 ETM+ 2012
Accuracy Assessment
Change Detection
10 3.4 Data collection and analysis
A questionnaire survey was conducted to fulfill the objectives and prove the study’s hypothesis.
The questionnaire survey was completed between February and March 2012. A semi-structured household-level questionnaire was developed to investigate the effects of natural disasters, which was pre-tested before conducting the survey. To clarify the purpose of the survey, informed consent was agreed upon with the respondents. A total of 144 household-level questionnaire surveys were conducted in three unions, namely Shyamnagar, Koikhali and Nildumur (Figure 3.1), with 55, 45 and 44 responses respectively. These three unions were selected because natural disasters are common in them, while the impact level of disasters is different. Non-probability random sampling procedures were followed for the household surveys (Bryman 2004). All the households were selected as the researcher walked through the roads of those areas.
Three focus group discussions (FGD) were conducted to investigate the reasons behind the land- use and land-cover changes that had occurred in the study area. In each union, one FGD has been conducted where both agricultural farmers and shrimp farmers were the main focus. The numbers of participants in the FGD were 9, 8 and 9 for each of the three unions respectively.
Four individual informal interviews were conducted through a semi-structured questionnaire;
these comprised two government and two non-government officials.
Some secondary data were collected in order to obtain an overview of the occurrence of disasters and their impact on the study area. These data were collected from the Internet, the Local Government Engineering Department, the Upazila Agriculture Extension Office, the Upzila Fisheries Office and the Meteorological Department of Dhaka, Bangladesh.
Frequency tests were utilized in order to compare categorical data, such as age, sex and occupation, across the three unions. Probit regression analysis was performed to prove the hypothesis of the study because all the variables collected from household survey showed normal distribution. The level of significance for the analysis was set to p<0.08.
3.5 Limitations
As the leaders of rural households in Bangladesh are male, the participation of female respondents was relatively less.
11 4. RESULTS
4.1 Land-cover change detection
Satellite images of Landsat TM and ETM+ were analyzed in order to achieve an overview of land-cover changes which had occurred between 1999 and 2012 in the study area. Figure 4.1 represents the true and false color Landsat satellite images of Shyamnagar Upazila and its surrounding areas. From these images, it can be seen that the changes that have occurred in land- cover are visible and separable. Further analysis was done by classifying these images using supervised classification.
Figure 4.1: False color images of Shyamnagar and its surrounding areas from year 1999 (up) and 2012 (down)
It is visible from the classified map that in the year 1999, dominant land-use practice on that area was paddy culture, represented by the color yellow. The color blue represents rivers, canals,
12
ponds and shrimp culture farms. Unused and built-up lands are represented by a pink color and vegetation by a green color, including mangrove forest and other vegetation (Figures 4.2 and 4.3).
Figure 4.2: Classified map of the study area from the year 1999
Figure 4.3: Classified map of the study area from the year 2012
13
The positive changes which have occurred in bare land and built-up areas and also the negative changes which have occurred in agriculture, water bodies and vegetation between 1999 and 2012. The majority of water bodies have been occupied by bare and built-up areas due to population pressures resulting in expansion of built-up areas. Some lands have been converted to bare lands due to the effects of water logging and salinity intrusion (Figure 4.4).
Figure 4.4: Change map bare land and built-up areas from1999-2012
Those changes occurring from water body to other land cover categories during 1999 to 2012 presented in Figure 4.5. Red represents those areas which converted to a bare land or built-up category from the water body category. Yellow represents those areas which converted to the agricultural land category, and green represents those areas where water body was converted to the vegetation category.
14
Figure 4.5: Water body change map 1999-2012
Figure 4.6 represents those negative changes occurring in agricultural land use from 1999 to 2012. Blue represents those areas which converted to the water body category from other categories of land-covers. Red and green represent those areas which converted to the bare land or built up category and to vegetation, respectively.
Areas which converted to other land cover categories from the vegetation category between the years 1999 to 2012 represented in Figure 4.7. Red represents those areas which converted to bare land from vegetation. Also, some vegetation land converted to agricultural land and water body are represented by yellow and blue, respectively.
15
Figure 4.6: Agriculture land change map from 1999-2012
Figure 4.7: Vegetation change map 1999-2012
16 4.1.1 Change detection Statistics
Chang detection statistics represents the percentage of differences in pixels between two classified images of the tear 1999 and 2012 (Table 4.1). The main changes occurred in bare land and built-up areas, which increased by almost 73 percent. Almost 48 percent of the agricultural land converted to other land cover categories within the last 13 years; 29 percent of agricultural land has been converted to the water body category and 23 percent of agricultural land converted to the bare land and built-up category. The second most significant change occurred in the water body category, which increased by almost 31 percent. Vegetation and tree species decreased by 3 percent within the last 13 years; 6 percent of vegetation has been converted to agricultural land and 4 percent to the water body category.
Table 4.1: Change detection Statistics from year 1999 to 2012
Categories
Percentage Built-up\bare
land
Water body Agriculture land
Vegetation Row Total
Class Total
Bare land\Built-up 31 5 23 4 100 100
Water body 47 93 29 4 100 100
Agriculture land 11 1 44 6 100 100
Vegetation 11 1 4 86 100 100
Class Total 100 100 100 100 0 0
Class Change 69 7 56 14 0 0
Image Difference 73 30 -48 -3 0 0
4.2 General characteristics of respondents
Baseline characteristics of the studied population are represented in the Table 4.2, where most of the respondents were male (83 percent), 62 percent of the population were in their middle age (25–30 years), and the major occupations were farming and fishing (17 percent in both cases).
Only 17 percent of total respondents had shrimp culture farms. Findings revealed that 90 percent of the total respondents were affected by disasters; where all households were in the Nildumur union, (100 percent)
17
Table 4.2: Baseline characteristics of the study population( N indicates number of respondents)
Area Shyamnagar
Union Koikhali Union
Nildumur Union
All
Characteristics N % N % N % %
Gender
Male 41 74 41 89 38 88 83
Female 14 26 5 11 5 12 17
Age
25-50 34 62 30 65 25 58 62
50-70 19 34 12 26 15 35 31
>70 2 4 4 9 3 7 7
Major Occupation
Business 7 13 3 7 10 23 14
Farmer 9 16 15 33 1 2 17
Fisherman 0 0 8 17 16 37 17
Housewife 6 11 5 11 4 9 10
Labor 7 13 6 13 3 7 11
Van Puller 7 13 0 0 0 0 5
Student, Teacher, Service Holder, Doctor, Retired
9 16 2 4 3 7 10
All Other 10 18 7 15 6 15 16
Family size
2 - 4 29 53 12 26 11 26 36
5 - 6 15 27 20 44 19 44 38
>6 11 20 14 30 13 30 26
Education among respondents (class)
0 6 11 4 9 5 12 11
1 -5 44 80 36 78 33 77 78
6 - 10 5 9 6 13 5 11 11
Shrimp Culture pond
Yes 5 9 5 11 14 33 17
No 50 91 41 89 29 67 83
Disaster Affected
Yes 42 76 44 96 43 100 90
No 10 18 2 4 0 0 8
N\A 3 6 0 0 0 0 2
4.3 Changes in land-use and their causes
Respondents from FGDs mentioned, about those major changes in the local practices and their reasons (Table 4.3). They mentioned that positive changes (increase) occurred in cropping
18
pattern, shrimp culture, deforestation, and urbanization. Whereas, negative changes (decrease) occurred in the paddy culture, dam or polders and drainage sector.
Table 4.3: Changes in local practices and their causes according to respondents Local practices Changes (pos+, Neg-) Reasons
Cropping pattern
++ Introduced new varieties of crops to combat with salinity problem and to increase overall production rate.
Shrimp culture ++++ Economically profitable
Paddy culture - - Salinity intrusion due to shrimp culture and Storm surge Deforestation ++ Storm surge and shrimp culture
Urbanization +++ Increasing population
Dam (polders) - Decreased due to river bank erosion and storm surge Land
degradation
++ Decreased fertility of land due to salinity, adulterant in fertilizers (up to 50 percent)
Drainage - Due to congestion and grabbing
Land status of the study area before shrimp cultivation is represented in Table 4.4. The analysis showed that a considerable portion which converted to shrimp culture pond was from agricultural land; this was mentioned by almost 20 percent of respondents. Conversion from forest, vegetation, and bare land to shrimp culture ponds was also mentioned by the respondents.
Table 4.4: Land status of study area before shrimp cultivation
Before Shrimp cultivation Frequency Percent (Multiple Value)
Forest 7 4.8
Agricultural land 29 20
Other (Bare land, Vegetation) 6 4.1
Don’t know 102 70
4.4 Factors responsible behind major land-use changes
Conversion of agriculture and other land-use to shrimp culture ponds were the main changes occurring in the study area. Respondents from FGDs mentioned that economic factors are the main drivers behind these changes (Table 4.5), as shrimp farming is more beneficial than paddy
19
culture. They also mentioned other factors behind significant changes in land-use, including social, political and natural changes.
Table 4.5: Factors responsible for land-use changes Factors Priority Description
Political 3 Land grabbing in terms of power and conversion to shrimp farm Economical 1 Shrimp culture is profitable comparing to paddy culture
Natural 4 Water logging due to storm surge and cyclone resulting in declining agriculture production; decrease of cattle and tree species.
Social 2 One person start shrimp culture in the paddy field at the beginning then others bound to convert their traditional practice to shrimp culture due to salinity intrusion
4.5 Disasters in the study area and their effects according to respondents 4.5.1 Salinity level of the study area
During the past few decades, the salinity level of the study area has increased significantly.
Salinity level of the Satkhira area, including Shyamnagar, where within the last 40 years salinity has increased by 3 percent (Table 4.6).
Table 4.6: Level of Salinity in study area
Salinity Level Salinity increase
over 4 decades Year S1
2-4 dS/m
S2
4.1-8 dS/m
S3
8.1-16 dS/m
S4
>16 dS/m
Area (000’ha) %
1973 26.50 85.60 34.50 10.90 4.76 3.02
2000 29.03 39.01 60.55 22.01
2009 31.00 32.96 69.72 28.58
Source: Miah, (2010). *S1 slightly saline, S2 Moderately Saline, S3 Saline and S4 Highly saline 4.5.2 Type and level of disasters on the study area
To identify the category of disasters occurring in the study area and their levels of impact, frequency tests of household data were utilized (Table 4.7). Almost 28 percent of respondents from the Nildumur union mentioned that cyclones affected them severely; almost 46 percent of
20
respondents from the Nildumur union mentioned tidal flood as a severe effect on them, while nearly 44 percent of respondents from the Koikhali union mentioned salinity as a severe problem in their area. The majority of respondents from the Shaymnagar and Koikhali unions (42 and 37 percent, respectively) mentioned water logging as severely affecting those areas.
Table 4.7: Type of disasters in study area
Disasters
Area Shaymnagar
Union
Koikhali Union
Nildumur Union
All
Level Frequency % Frequency % Frequency % %
Cyclone
Severe 10 18 3 7 12 28 18
High 10 18 5 11 17 40 22
Moderate 9 16 14 30 11 26 24
Low 8 15 11 24 2 4 15
No Impact 18 33 13 28 1 2 21
Tidal flood
Severe 10 18 5 11 20 47 24
High 19 35 20 44 11 25 35
Moderate 17 31 12 25 9 21 26
Low 2 4 6 13 2 5 7
No Impact 7 12 3 7 1 2 8
Salinity
Severe 9 16 20 44 7 16 25
High 11 20 12 26 10 23 23
Moderate 5 9 4 9 17 40 18
Low 5 9 4 9 2 5 8
No Impact 25 46 6 12 7 16 26
Water logging
Severe 23 42 17 38 3 7 30
High 9 16 7 15 4 9 14
Moderate 8 15 7 15 3 7 13
Low 5 9 7 15 21 49 23
No Impact 10 18 8 17 12 28 20
21
4.5.3 Impacts of disasters on agricultural production, income and migration
Since the majority of people in the study area depend upon agricultural activities, probit regression analysis was used to test the impact of disasters on agricultural production (Table 4.8).
The analysis showed that the probability of agricultural production decreased significantly (10%
level) for agricultural farmers compared to other occupation. Further, the probability of agricultural production decreased significantly (1% level) due to cyclones and tidal flooding.
Table 4.8: Effect of disasters on agricultural production
Agricultural production (decreased 1, increased 0) Coef. Std. Err. z P>z
Age (years) 0.01 0.01 0.88 0.38
Gender (male 1, female 0) -0.02 0.40 -0.06 0.95
Family size (number) 0.13 0.06 2.1 0.04
Occupation farmers (1 if agricultural farmer, otherwise 0) 0.64 0.37 1.76 0.08 Occupation service and business (1 if service and business,
otherwise 0)
0.32 0.34 0.94 0.35
Cyclone(1 if have any negative impact, otherwise 0) 0.85 0.24 3.56 0.00 Tidal flood (1 if have any negative impact, otherwise 0) 0.86 0.32 2.65 0.01 Salinity (1 if have any negative impact, otherwise 0) 0.22 0.25 0.89 0.37
Constant -2.63 0.73 -3.63 0.00
The probability of overall income of the respondents decreased significantly (10% level) for agricultural farmers compared to other occupation (Table 4.9). The probability of overall income has also decreased significantly because of cyclones and tidal surges at the 1 and 4 percent levels, respectively.
The probability of outward migration increased significantly (10% level) for farmers. Further, migration probability increased due to tidal floods significantly at a 5 percent level (Table 4.10).
22
Table 4.9: Effect of disasters on income of the respondents
Overall Income (decreased 1, increased 0) Coef. Std. Err. z P>z
Age (years) 0.01 0.01 1.32 0.19
Gender (male 1, female 0) -0.72 0.40 -1.81 0.07
Family size (number) -0.01 0.06 -0.26 0.79
Occupation Farmers (1 if agricultural farmer, otherwise 0) 0.65 0.36 1.8 0.07 Service and business (1 if service and business,
otherwise 0)
0.37 0.34 1.07 0.28
Cyclone (1 if have any negative impact, otherwise 0) 0.68 0.23 2.89 0.00 Tidal flood (1 if have any negative impact, otherwise 0) 0.64 0.31 2.05 0.04 Salinity (1 if have any negative impact, otherwise 0) 0.25 0.24 1.03 0.30
Constant -1.17 0.67 -1.76 0.08
Table 4.10: Effects of disasters on human outwards migration
Outward migration (yes 1, no 0) Coef. Std. Err. z P>|z
Age (years) -0.01 0.01 -0.56 0.58
Gender (male 1, female 0) -0.51 0.55 -0.94 0.35
Family size (number) 0.15 0.06 2.26 0.02
Occupation Farmers (1 if agricultural farmer, otherwise 0) 0.94 0.51 1.84 0.07 Service and business (1 if service and business,
otherwise 0)
0.79 0.49 1.62 0.10
Cyclone (1 if have any negative impact, otherwise 0) 0.47 0.29 1.63 0.10 Tidal flood (1 if have any negative impact, otherwise 0) 0.99 0.50 2.00 0.04 Salinity (1 if have any negative impact, otherwise 0) -0.18 0.28 -0.63 0.53
Constant -2.76 0.82 -3.36 0.00
23 4.5.4 Consequential effects of disasters
Respondents from FGD’s mentioned that both cyclone and tidal surges caused devastating effects, which ultimately led people to change their place of residence and their occupation. All other disasters, including drought, irregular rainfall, salinity and water logging have had similar effects and as a result, caused changes in the overall livelihood pattern (Table 4.11).
Table 4.11: Disasters and their consequential effect on the study area Disasters Occurrences
in last 20 years
Effects Adaptation
strategies
Remark
Cyclone\Tidal flood
2 times Destruction of house and properties
Destruction of paddy and other agricultural crops
Cattle and fish decrease
Taking loan from bank or microcredit NGO’s and migration to other places
From 2009 to 2011 there was no production of crops in the field, most of the people were involve in cutting soil Drought From 2001
to present, every year in summer
Scarcity of irrigation and drinking water
Damages of crops
Death of shrimp fry and growth of shrimp reduced
Nothing Nothing
Irregular rainfall
1 time excessive rainfall
Damage of crops for excessive rainfall
Water logging
- Nothing Nothing
Salinity Increasing from the year 2000
Crop production reduced
Shrimp growth hampered
Reduction of native tree species
New verities (Hybrid) crop cultivation
Nothing
Water logging Occurring occasionally from the year 2000
Damage to the
agricultural crops
Death of cattle and destruction of trees
Loss in fish culture
Nothing Due to drainage congestion
24 4.5.5 Diseases outbreak after disasters
Respondents mentioned about various disease outbreaks after the latest cyclone Aila in 2009.
Most of the respondents (91 and 95 percent) were affected by diarrhea and headache, and 92 percent of the respondents were affected by dysentery, pneumonia and jaundice. Skin diseases, typhoid fever and viral diseases of the eye have been experienced by almost 67, 57 and 44 percent respondents, respectively (Table 4.12).
Table 4.12: Diseases outbreak after latest disaster (Cyclone Aila 2009)
Diseases After (Aila) Frequency Percent (Multiple Value)
Diarrhea 132 91
Cholera 12 8
Skin Diseases 98 68
Chicken pox 18 12
Typhoid /Fever 83 57
Headache 138 95
Viral diseases of eye 64 44
Dysentery, Jaundice and Pneumonia 133 92
4.6 Migration in the study area and their causes
About 21 percent of the respondents mentioned they had migrated within the last 13 years. In Shyamnagar union, 96 percent of the respondents mentioned that there was no outward migration from 1999 to 2012 (Table 4.13). The principal causes behind outward migration (Table 4.14) are job and income, as mentioned by almost 16 and 13 percent of the respondents, respectively.
25 Table 4.13: Outward migration in study area
Area Shyamnagr
Union
Koikhali Union
Nildumur Union
All
Outward migration N % N % N % %
Yes 2 4 8 17 19 44 21
No 53 96 38 83 24 56 79
Table 4.14: Causes of outward migration in study area
Causes of Migration Frequency Percent (Multiple Value)
Job 23 15.9
Income 19 13.1
Natural Disasters 15 10.3
Others (River Bank Erosion, Better Life and Education) 17 11.7
No Migration 92 63.4
4.7 Changes in tree and fish species
Most respondents (93 and 92 percent) from the household survey mentioned that tree and fresh water fish species decreased between1999 and 2012 (Table 4.15). The main reason behind these changes was salinity intrusion and occurrences of cyclone, as mentioned by respondents from the FGD.
Fish species became extinct from the study area within last 13 years are Rui (Labeo rohita), Katla (Catla catla), Climbing perch (Anabas testudineus), Snakehead murrel (Channastriata), Walking catfish (Clarias batrachus), Puntio barb (Puntius puntio), Spotted snakehead (Channa punctata), Wallago (Wallagoattu), Tengra (Batasio batasio).
Type of tree species became extinct from the study area are Mango (Mangiferaindica), Lime (Cytrusaurantifolia), Jackfruit (Artocarpus heterophyllus), Coconut (Cocos nucifera), Papaya (Carica papaya), Guava (Psidiumguajava), Betel nut (Areca catechu), Palmira palm (Borassus flabellifer), Sapota (Manilkara achras), Tamarind (Tamarindus indica), Babla (Acacia nilotica),
26
Jamun (Syzygium cumini), Indian lilac (Azadirachta indica), Mahogany (Swietenia macrophylla), and most of the flower trees.
Table 4.15: Changes in tree and fish species from 1999 to 2012 (according to respondents) in study area
Category Tree species (%) Fish species (%)
Increased 3 4
Decreased 93 92
No change 4 4
27 5. DISCUSSION
5.1 Changes in land-use and land-cover in the study area from 1999 to 2012
Changes area prominent feature in the land-use and land-cover of any social-ecological system.
It is not possible to keep any social-ecological system constantly unchanged. Humans are changing the land-cover continuously to cope with the changing environment. Land-use has been changed rapidly and drastically all over the world over the last few decades (Xing et al. 2006).
From the satellite image analysis, it is also evident that very significant land-cover change (positive and negative) has occurred in the study area within the last thirteen years. Increases have occurred in the water bodies category because shrimp culture ponds have increased.
Increases have also occurred in the built-up and bare land category. Built up areas increased due to an increase in population, but this does not explain increases in the bare land category. Bare land might increase due to the secondary effects of various disasters. For example, water logging occurs in the study area due to tidal surges and floods. Because it is surrounded by coastal embankments, the natural drainage system is blocked due to dams or polders, land grabbing, siltation and decreases in upstream water flow and, as a result, salinity intrusion occurs in agricultural land and other areas. Thus, most of the lands of the area have become unproductive for any use due to long-term inundation by saline water resulting in increases in the bare land category. UN-HABITAT reported that because of high intensity disasters, large parts of an area may become uninhabitable due to long-term inundation (UN-HABITAT 2010).
Decreases occurred in the agricultural land category due to an increase in shrimp culture ponds in the area. The FGDs showed that agricultural land has been converted to shrimp culture ponds for economic, social and political reasons. Decreases have also occurred in the vegetation category, including trees and mangrove forest. This may be due to occurrences of both natural disasters and other anthropogenic activities. Respondents from the household survey also mentioned that tree species have decreased as a result of cyclones. Salinity intrusion due to shrimp farms is another reason for destruction of trees, as the forest is cleared to cultivate shrimp (Haque et al. 2008).
Respondents from the FGDs mentioned other land-use changes in the study area, including increased urbanization and land degradation. Urbanization increased due to population pressure, while land fertility and plant production yields both decreased due to increased salinity levels.
